See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/264431350

Fast transmission from the dopaminergic ventral midbrain to the sensory

cortex of awake primates

Article  in  Brain Structure and Function · August 2014

DOI: 10.1007/s00429-014-0855-0 · Source: PubMed

CITATIONS

5READS

202

6 authors, including:

Some of the authors of this publication are also working on these related projects:

Psychophysiology of auditory perception View project

Quantitative mapping of cytochrome oxidase and energy metabolism in the brain View project

Judith Mylius

German Primate Center

12 PUBLICATIONS   58 CITATIONS   

SEE PROFILE

Max F.K. Happel

Leibniz Institute for Neurobiology

40 PUBLICATIONS   321 CITATIONS   

SEE PROFILE

Alexander Georgievich Gorkin

Russian Academy of Sciences

32 PUBLICATIONS   106 CITATIONS   

SEE PROFILE

Ying Huang

Leibniz Institute for Neurobiology

51 PUBLICATIONS   232 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Judith Mylius on 19 September 2014.

The user has requested enhancement of the downloaded file.

ORIGINAL ARTICLE

Fast transmission from the dopaminergic ventral midbrainto the sensory cortex of awake primates

Judith Mylius • Max F. K. Happel •

Alexander G. Gorkin • Ying Huang •

Henning Scheich • Michael Brosch

Received: 20 March 2014 / Accepted: 21 July 2014

� Springer-Verlag Berlin Heidelberg 2014

Abstract Motivated by the increasing evidence that

auditory cortex is under control of dopaminergic cell

structures of the ventral midbrain, we studied how the

ventral tegmental area and substantia nigra affect neuronal

activity in auditory cortex. We electrically stimulated 567

deep brain sites in total within and in the vicinity of the two

dopaminergic ventral midbrain structures and at the same

time, recorded local field potentials and neuronal dis-

charges in cortex. In experiments conducted on three

awake macaque monkeys, we found that electrical stimu-

lation of the dopaminergic ventral midbrain resulted in

short-latency (*35 ms) phasic activations in all cortical

layers of auditory cortex. We were also able to demonstrate

similar activations in secondary somatosensory cortex and

superior temporal polysensory cortex. The electrically

evoked responses in these parts of sensory cortex were

similar to those previously described for prefrontal cortex.

Moreover, these phasic responses could be reversibly

altered by the dopamine D1-receptor antagonist SCH23390

for several tens of minutes. Thus, we speculate that the

dopaminergic ventral midbrain exerts a temporally precise,

phasic influence on sensory cortex using fast-acting non-

dopaminergic transmitters and that their effects are mod-

ulated by dopamine on a longer timescale. Our findings

suggest that some of the information carried by the neu-

ronal discharges in the dopaminergic ventral midbrain,

such as the motivational value or the motivational salience,

is transmitted to auditory cortex and other parts of sensory

cortex. The mesocortical pathway may thus contribute to

the representation of non-auditory events in the auditory

cortex and to its associative functions.

Keywords Deep brain stimulation � Primate � Auditory

cortex � Glutamate corelease � Ventral tegmental area

Introduction

There is converging evidence that auditory cortex is under

control of dopaminergic cell structures in the ventral

midbrain. Involvement of the dopaminergic ventral mid-

brain is suggested by the finding that neurons in auditory

cortex reflect rewards given during the performance of

auditory tasks and reward prediction errors (Brosch et al.

2011a). Reward-related activations are generally consid-

ered to be associated with activation of dopaminergic

neurons in the ventral tegmental area and substantia nigra

(Schultz 1998, 1997). In addition, auditory cortex is

recruited during the performance of auditory working

memory tasks (Fritz et al. 2005; Huang et al. 2012), a

function generally thought to involve dopaminergic

J. Mylius (&) � Y. Huang � M. Brosch

Special Laboratory Primate Neurobiology, Leibniz Institute for

Neurobiology, Brenneckestraße 6, 39118 Magdeburg, Germany

e-mail: Judith.Mylius@lin-magdeburg.de

M. F. K. Happel

Department Systems Physiology of Learning, Leibniz Institute

for Neurobiology, Brenneckestraße 6, 39118 Magdeburg,

Germany

A. G. Gorkin

Institute of Psychology, Russian Academy of Sciences,

Yaroslavskaya Street 13, 129366 Moscow, Russia

H. Scheich

Emeritus Group Lifelong Learning, Leibniz Institute for

Neurobiology, Brenneckestraße 6, 39118 Magdeburg, Germany

H. Scheich � M. Brosch

Center for Behavioral Brain Sciences, Otto-von-Guericke-

University, Universitatsplatz 2, 39106 Magdeburg, Germany

123

Brain Struct Funct

DOI 10.1007/s00429-014-0855-0

mechanisms in the prefrontal cortex (Sawaguchi and

Goldman-Rakic 1994; Chudasama and Robbins 2004). An

fMRI study found that subjects had increased BOLD sig-

nals in auditory cortex in response to tones after oral

application of L-dopa when they were engaged in an

instrumental learning task (Weis et al. 2012). A brain slice

study revealed that dopamine can modulate glutamate

release in auditory cortex (Atzori et al. 2005). Dopamine

could also influence auditory cortex indirectly through the

inferior colliculus (Gittelman et al. 2013). Similar short-

term effects of dopamine have also been demonstrated in

visual cortex (Arsenault et al. 2013). In addition to these

short-lasting effects, the dopaminergic ventral midbrain

can also have long lasting effects on neuronal activity in

auditory cortex, resulting in permanent structural changes

and underlying learning (Stark and Scheich 1997; Schic-

knick et al. 2008, 2012; Kudoh and Shibuki 2006; Bao

et al. 2001, 2003).

Dopaminergic control of auditory cortex may arise from

direct and indirect inputs from the ventral tegmental area

and the substantia nigra (Scheibner and Tork 1987;

Campbell et al. 1987; Budinger et al. 2008), and in mon-

keys also from the retrorubral field (Gaspar et al. 1992;

Williams and Goldman-Rakic 1998). In accordance with

the rostro-caudal density gradient in cerebral cortex dopa-

minergic terminals and receptors show highest density in

rostral regions of primate auditory cortex and their density

falls off at more caudal regions (Campbell et al. 1987;

Lewis et al. 1987; Lidow et al. 1991; Zilles et al. 2002;

Amunts et al. 2012). At the same time, dopaminergic ter-

minals and receptors have a layer-specific pattern in pri-

mate and human auditory cortex, with highest terminal

density in layer I and in infragranular layers and highest

D1-receptor density in supragranular layers (Campbell

et al. 1987; Lewis et al. 1987; Lidow et al. 1991; Zilles

et al. 2002; Amunts et al. 2012). Direct influences of the

dopaminergic ventral midbrain on auditory cortex may also

be exerted by glutamate corelease from the dopaminergic

terminals (Lavin et al. 2005; Yamaguchi et al. 2011) as

well as by projections of glutamatergic and GABAergic

cells, which are interspersed with the dopaminergic neu-

rons and can comprise up to 40 % of the cells in the ventral

tegmental area (Margolis et al. 2006; Fields et al. 2007).

Little is known about how the projections of the dopa-

minergic ventral midbrain affect neuronal activity in

auditory cortex (Atzori et al. 2005). Further insight into this

question may be garnered from the studies in rodents that

have been performed on the main target of the mesocortical

pathway, namely the prefrontal cortex, and on the motor

cortex. These studies have revealed that stimulation of the

dopaminergic ventral midbrain or intracortical dopamine

application can result in phasic (\&1 s) or tonic ([&1 s)

changes of cortical neuronal activity.

Brief electrical stimulation of the ventral tegmental area

and substantia nigra phasically increases or decreases

neuronal activity in prefrontal cortex for up to a few hun-

dred milliseconds. On the cellular level, initially an EPSP

with a time constant of a few tens of milliseconds is evoked

about 10 ms after stimulation (Bernardi et al. 1982; Mer-

curi et al. 1985; Lavin et al. 2005), indicating conduction

velocities of the mesocortical projection system in the

order of 10 m/s (Deniau et al. 1980; Thierry et al. 1980). In

many neurons, the EPSP is followed by an IPSP that lasts a

few hundred milliseconds. In some cortical neurons, the

postsynaptic potentials result in phasic decreases and

increases of their firing (Au Young et al. 1999; Ferron et al.

1984; Mantz et al. 1988; Jay et al. 1995; Pirot et al. 1992;

Thierry et al. 1992). These observations in single neurons

are in good agreement with the results of electrophysio-

logical and optical recordings of cortical population

activities (Lavin et al. 2005; Watanabe et al. 2009).

Pharmacological manipulations have revealed that the

phasic cortical responses to electrical stimulation are not

mediated by intracortical D1- and D2-receptors but rather

are generated by GABA and glutamate. These transmitters

are released by the mesocortical GABAergic (Carr and

Sesack 2000) and glutamatergic pathway (Yamaguchi et al.

2011), or the latter coreleased from dopaminergic fibers

(Lavin et al. 2005; Yamaguchi et al. 2011). The dopamine

released after activation of the dopaminergic ventral mid-

brain still affects cerebral cortex, although in a different

manner and for a longer time (Sawaguchi and Matsumura

1985; Sawaguchi et al. 1986; Matsumura et al. 1990; Jacob

et al. 2013; Hosp et al. 2011; Happel et al. 2014).

In the current study, we aimed to analyze the phasic

effects of activation of the dopaminergic ventral midbrain

on neuronal activity in the auditory cortex of primates, i.e.,

in an animal model whose mesocortical dopaminergic

system is more similar to that of humans than that of

rodents (Berger et al. 1991; Smith et al. 2013). To this end,

we adopted the paradigm that has been widely used in

studying neural transmission from the dopaminergic ven-

tral midbrain to prefrontal and motor cortex in rodents. We

applied brief electrical stimulation trains to the dopami-

nergic ventral midbrain and measured neuronal activity in

auditory cortex. With this approach, we addressed the

following questions: does electrical stimulation in the

dopaminergic ventral midbrain affect local field potentials

(LFPs) and neuronal discharges in auditory cortex? Is the

most densely expressed dopamine D1-receptor involved in

these effects (Zilles et al. 2002; Amunts et al. 2012)? Are

there differences between cortical layers, reflecting the

layer-specific pattern of dopaminergic terminals and

receptors (Campbell et al. 1987; Zilles et al. 2002; Amunts

et al. 2012)? What are the auditory response properties of

the neurons whose action potentials can be modified by

Brain Struct Funct

123

electrical stimulation of the dopaminergic ventral mid-

brain? Our experimental approach also enabled us to

examine effects of dopaminergic ventral midbrain stimu-

lation on cortex subserving other sensory modalities, i.e.,

the somatosensory cortex and the superior temporal poly-

sensory cortex. The current experiments were conducted on

awake animals to exclude the effects of anesthesia on the

cortical state. Tonic effects of activation of the dopami-

nergic ventral midbrain on neuronal activity in the auditory

cortex will be presented elsewhere.

Methods

Subjects

Experiments were conducted on three adult male Old

World cynomolgus monkeys (Macaca fascicularis; mon-

keys W, I, E). They had been previously used in other

studies in which they performed a behavioral task using

positive reinforcement and in which neuronal activity was

recorded from their auditory cortex. All experiments were

carried out under approval of the animal care and ethics

committee of the State of Sachsen-Anhalt (No. 28-14

42502/2-806 IfN) and in accordance with the guidelines for

animal experimentation of the European Communities

Council Directive (86/609/EEC).

Animal preparation

Monkeys were surgically implanted with a head holder into

the skull to allow atraumatic head fixation (Brosch and

Scheich 2008). A recording chamber (21 mm diameter)

was implanted over the left (monkey W and I) or the right

temporal cortex (monkey E). It was centered on Horsley–

Clarke coordinates A7.5 to A8.5 and D8.5 to D12.5 (con-

ventions as in Szabo and Cowan, 1984). For all surgical

procedures, monkeys were kept under deep general anes-

thesia with a mixture of ketamine (2 mg/kg) and xylazine

(5 mg/kg). Surgery was followed by a full treatment with

antibiotics (enrofloxacin, 0.2 ml/kg im) and analgesics

(carprofen, 0.1 ml/kg) for up to 5 days. The dura mater was

treated daily with the local disinfectant and anti-inflam-

matory agent Dexamytrex (Mann).

Electrophysiology

Experiments were carried out in an electrically shielded,

sound-attenuated double-walled room (1202-A, IAC) that

was illuminated with an incandescent light bulb. During the

experiments the head-restrained monkeys sat quietly and

were awake for several hours in a primate chair. They were

monitored by a video camera (ICD-848P, Ikegami) and

recorded by a computer (Pinnacle Studio 10, Pinnacle).

With the aid of two separate manipulators, we inserted

stimulation and recording microelectrodes through the

same recording chamber into the brain (Fig. 1a).

Neuronal activity in primary auditory cortex and in

adjacent parts of the overlying secondary somatosensory

cortex and the underlying superior temporal polysensory

Fig. 1 Experimental approach. a Schematic view of a monkey with a

head holder for head restraint and a recording chamber providing

access to various deep brain structures and the cortex. Red and blue

lines indicate the approximate orientations of stimulation and

recording electrodes, respectively. b Coronal section through a

monkey brain at the level of the dopaminergic ventral midbrain and

the auditory cortex (see inset). The left half of the section shows an

anatomical magnetic resonance (MR) image of monkey I. MR

contrast was optimized to highlight iron-containing brain nuclei, such

as the substantia nigra (SN), indicated by the encircled region, and to

highlight iron deposits. The white arrowhead points to iron deposits

along an example electrode track covering the ventral tegmental area

(VTA) and the contralateral SN. The right half of the section shows a

corresponding anatomical schematic of cortical gray and white

matter, ventricles (black), and various subcortical brain nuclei. We

also highlighted other brain regions of interest for the current study,

such as auditory cortex (AC), secondary somatosensory cortex (S2),

superior temporal polysensory cortex (STP), lateral sulcus (ls), and

superior temporal sulcus (sts). Scale bar 5 mm

Brain Struct Funct

123

cortex was recorded with a sixteen-channel microelectrode

system (System Eckhorn, Thomas Recording), equipped

with 5–10 tungsten microelectrodes (80 lm diameter,

2–2.5 MX). During most experiments, the electrodes were

arranged in a 4-by-4 grid with a lateral spacing of 305 lm;

sometimes electrodes were arranged as a bundle in a can-

nula. The electrode system was oriented slightly off the

dorsoventral plane (10�). Electrodes entered the cortex

between stereotactic coordinates A5 to A9 and D7.5 to

D12.5 and subsequently were moved through the somato-

sensory cortex into the auditory cortex, and sometimes also

into the superior temporal polysensory cortex (Fig. 1b).

After preamplification the signal on each electrode was

amplified and bandpass filtered (LFP-01, Thomas Record-

ing; MCP-Plus, Alpha-Omega Engineering) to extract

action potentials (0.5–5 kHz) and local field potentials

(LFP; 0.1–100 Hz in monkey W, 1–150 Hz in monkeys I

and E). All signals were fed into an AD data acquisition

system (32-channel BrainWave, DataWave Technologies,

32-channel Alpha-Map, Alpha-Omega, 32-channel Chee-

tah, Neuralynx for monkeys W, I and E, respectively).

Sampling rates for LFPs ranged between 640 and 659 Hz;

for action potentials they were between 30 and 42 kHz. By

means of built-in spike detection tools of the data acqui-

sition systems (threshold crossing and spike duration), the

action potentials of the somas of a few neurons in the

vicinity of the electrode tip were discriminated (Super and

Roelfsema 2005). Only the time stamps and waveforms of

the action potentials were stored. From multiunit records

with electrically evoked responses we extracted the spikes

of single units off-line by means of principal component

analysis.

Electrical stimulation

For electrical stimulation a separate seven-channel micro-

electrode system was used (System Eckhorn, Thomas

Recording). It was equipped with tungsten microelectrodes

(80 lm diameter) with impedances of 0.1–0.4 MX, which

also allowed recording of neuronal activity. One of the

electrodes was connected to a single-channel constant

current stimulator (STG 1001, Multichannel Systems). We

applied monopolar stimulation (referenced to the bolts of

the implanted headholder) that consisted of short trains

with six rectangular, biphasic pulses (positive first, 350 ls

phase duration, no pause) at 1,000 Hz with an amplitude of

100 lA. These trains were repeated 50–100 times every

900 ms. During an experimental session, 50–150 of such

stimulation blocks were applied to the monkeys. Our

stimulations were likely strong enough to excite most

cortical sites, because during occasional tests, we found

robust neuronal responses already with trains consisting of

fewer pulses (3), smaller pulse frequencies (400 Hz),

smaller currents (50 lA), and higher repetition rates (5/s).

Electrode trajectories were planned based on brain

atlases (Szabo and Cowan 1984; Martin and Bowden

1996), which were refined by anatomical magnetic reso-

nance (MR) scans of the individual monkey brains (Fig. 1b

for monkey I). MR images were acquired on a 7T Scanner

(Magnetom, Siemens) equipped with a head coil (Nova

Medical, Inc.). Gradient echo scans (TR 266 ms, TE

15 ms, flip angle 30�, averages 2, FOV 140 9 105 mm,

matrix 640 9 420, slice thickness 1.0 mm, interslice gap

20 %) in all three planes were optimized to highlight iron-

containing brain nuclei that were of interest for the current

study, such as the substantia nigra and red nucleus.

Electrodes were inserted into the brain at an angle of 28�from the horizontal plane between stereotactic coordinates

A10–A14 and D4–D7.5 and advanced up to 35 mm

(Fig. 1b). In a single monkey (I) subsequent to the last

recordings small iron deposits were made in the dopami-

nergic ventral midbrain with anodal, monophasic, direct

current pulses (10 lA, 90 s) from the tip of an iron

microelectrode (80 lm diameter, 0.3–0.6 MX). After ter-

mination of experiments the head holder of monkey I was

explanted, enabling MR imaging of the iron deposits to

enable electrode track reconstruction (white arrowhead in

Fig. 1b).

After an initial exploration phase during which we

inserted electrodes at different angles into different depths

of the brain, we were able to reproducibly place electrodes

at the desired depths and with the desired physiological

characteristics. No histological verification of stimulation

sites was undertaken because in each monkey more than

100 stimulation sites were tested and all 3 monkeys of the

current study are still being used for additional studies, in

accordance to the principles of Replacement, Reduction,

and Refinement for animal experimentation. In addition,

we used the following criteria to estimate whether the

electrode was within the dopaminergic ventral midbrain or

within some other deep brain structure.

1. We visually and auditorily monitored the spiking

activity along the entire electrode track, with a special

emphasis on identifying spiking that is considered to

be specific for midbrain dopaminergic cells, particu-

larly for the substantia nigra pars compacta, such as

slow tonic firing (1–10 Hz) or phasic burst firing

(14–20 Hz) with characteristic bi- or triphasic action

potentials of long durations greater than 2 ms (Guyenet

and Aghajanian 1978; Grace and Bunney 1980, 1983,

1984a, b; Freeman et al. 1985; Freeman and Bunney

1987; Chiodo 1988; Kiyatkin 1988; Kiyatkin and

Zhukov 1988; Diana et al. 1989; Grace and Onn 1989;

Grace 1991).

Brain Struct Funct

123

2. Some brain structures surrounding the dopaminergic

ventral midbrain can be identified by their specific

physiological characteristics. Of particular help was the

red nucleus which, for many of our electrode trajectories

(e.g., Fig. 3a), is located immediately before the ventral

tegmental area and always mediodorsal from the

substantia nigra. We identified the red nucleus by

(I) its specific movement-related discharges (Gibson

et al. 1985), (II) its characteristic electrically evoked

movements, such as movements of the contralateral

proximal extremities, shoulders, and upper arms, closing

of eyelids, pulling back the mouth angle, flattening the

cheek, or drawing back the pinna (Delgado 1965; Larsen

and Yumiya 1980; Gibson et al. 1985; Kennedy et al.

1986; Houk et al. 1988; Lovell et al. 2014), and (III) its

responsiveness to acoustical stimulation (Massion and

Albe-Fessard 1963; Irvine 1980; Lovell et al. 2014).

3. At selected locations, we tested whether the averaged

electrically evoked field potential was modified by the

dopamine D1-receptor antagonist SCH23390 (see

below).

4. We evaluated the shape of the electrically evoked field

potential in auditory cortex that was evoked by

electrically stimulating a deep brain structure and

assessed whether it matched that of the sites whose

electrically evoked field potential could be modified by

SCH23390.

Acoustical stimulation

Pure tones were generated by a computer, interfaced with

an array processor (AP2-card, Tucker-Davis Technologies)

at a sampling rate of 100 kHz, and D/A converted (DA1,

Tucker-Davis Technologies). Tones were presented at 40

different frequencies, each repeated 10 times. Frequencies

were equidistantly spaced on a logarithmic scale over a

range of about eight octaves (0.11–27.2 kHz), covering the

spectral sensitivity of the neurons under consideration.

Tones were presented in a pseudorandom order with a

constant interonset interval of 500 or 900 ms. Tone dura-

tion was 100 ms, including 5-ms ramps. In addition, we

tested the neurons’ auditory responsiveness with brief

clicks (5 ms duration, 1-ms ramps; WG1, Tucker-Davis

Technologies), which were repeated 50–150 times with a

constant interonset interval of 900 ms. All signals were

attenuated by a programmable attenuator (PA4, Tucker-

Davis Technologies), amplified (Pioneer, A204) and cou-

pled to a free-field loudspeaker (Manger), located

approximately 1.2 m from the animal. Sounds were pre-

sented at approximately 65 dB SPL. The sound pressure

level was measured with a free-field 1/2’ microphone

(40AC, G.R.A.S.) located near to the monkey’s head.

Pharmacology

At selected sites, we tested whether deep brain stimulation

activated the mesocortical pathway from the dopaminergic

ventral midbrain to auditory cortex by either intramuscular

or intracortical injections of the dopamine D1-receptor

antagonist SCH23390, i.e., of the antagonist that targeted

the most densely expressed dopamine receptor in primate

and human auditory cortex (Zilles et al. 2002; Amunts et al.

2012). The intracortical injections served to test in one

monkey (E) whether we could reproduce the effects of

intramuscular injections and thus to show that SCH23390

had effects on the receptors within the auditory cortex.

For systemic applications, the monkeys received an

intramuscular bolus injection of 0.03 mg/kg SCH23390

diluted in 0.1 ml/kg physiological saline in the occiput. For

intracortical applications, we used different doses of

SCH23390, diluted in physiological saline: 0.3 ll of a

50 mM or 2.4–3 ll of a 55, 62, 83, and 93 solutions. Using

a Hamilton syringe, we injected these solutions over a

period of 90 s via a micropipette (Thomas Recording) that

was brought into the vicinity (less than 1 mm) of the

recording electrodes using the same 16-channel microma-

nipulator that was used for the recordings from auditory

cortex. Most of our intracortical injections were sufficient

to affect all recording sites because a 3-ll injection of a

SCH23390 solution has been shown to spread about 3 mm

into the cerebral tissue (Sawaguchi and Goldman-Rakic

1991).

To examine the time course of the effects of the

antagonist on the results of deep brain stimulation, the

following protocol was used. (1) We measured the prein-

jection activity in auditory cortex that was evoked by deep

brain stimulation. (2) Within the following 5 min, the

antagonist was injected. Subsequently, the electrically

evoked activity was repetitively measured approximately

every 5–10 min for 1–3 h, until recovery.

Control experiments included (1) electrical stimulation

only without injection and (2) electrical stimulation com-

bined with systemic injection of 0.9 % physiological saline

instead of SCH23390.

Data analysis

Custom written MATLAB (version 2007b, MathWorks)

programs were used for off-line analyses.

Analyses of local field potentials

Evoked potentials in cortex

For each stimulation/recording pair, we averaged the LFPs

recorded in sensory cortex relative to the onset of electrical

Brain Struct Funct

123

stimulation (C50 trials). Because the stimulation generated

an artifact in the LFP records that lasted at the longest

20 ms, the initial portion of the electrically evoked

potential (EEP) could not be considered in this study and,

therefore, is never included in any of the EEPs shown in

this article (e.g., Fig. 2a). To compensate for the default

settings of the different filter systems used in the three

monkeys, all EEPs were band-pass filtered between 1 and

45 Hz, using a zero-phase forward and reverse, acausal,

second-order digital Butterworth filter.

For further purposes, we characterized each EEP by two

measures. (1) We calculated the average root mean square

(RMS) of the EEP value within the temporal interval of

20–170 ms after onset of the stimulation train. We

restricted our analysis to this interval because it contained

the most prominent and reproducible deflections of the

Fig. 2 Electrical stimulation of the dopaminergic ventral midbrain

evokes a polyphasic field potential in auditory cortex that is under

dopamine D1-receptor control. a Electrically evoked potentials

(EEPs) are modified after intramuscular injection of the D1-receptor

antagonist SCH23390. In the heat plot, rows show color-coded

amplitudes of the EEPs that were repeatedly measured after injection

(red indicating positive deflections, blue indicating negative deflec-

tions). Note that from 15 to 45 min after injection EEPs were

different from that before injection, whereas starting from 50 min

after injection the EEP was recovered. The three black lines show the

EEPs 5, 25 and 85 min after injection. The changes of the EEP were

quantitatively determined by calculating the Pearson correlation

between the EEP before injection with those at different times after

injection (right part of a). The red dashed line denotes the correlation

coefficient of 0.6 that was used to find time points at which an EEP

was modified. b Heat plot of the EEPs that were repeatedly measured

without any pharmacological intervention (control condition). Note

that the EEP was highly reproducible over the entire observation

interval and that Pearson correlations were always above 0.9. c Time

course of the effect of the D1-receptor antagonist SCH23390 on EEPs

in auditory cortex. The plot shows, for different times after injection,

the percentage of sites in the dopaminergic ventral midbrain from

where an EEP could be elicited that was significantly modified by the

antagonist. Blue and red curves indicate intramuscular and intracor-

tical injections, respectively. Insets show data of the individual

monkeys. d Template EEP characteristic for the dopaminergic ventral

midbrain. It was obtained by averaging all EEPs that were signifi-

cantly modified by the D1-receptor antagonist SCH23390 (see text).

Gray shading indicates s.e.m. Note that the initial 20 ms after onset of

stimulation are omitted because of the stimulus artifact

Brain Struct Funct

123

EEP. (2) We determined the similarity of an EEP to a

template EEP that was characteristic of the dopaminergic

ventral midbrain. To this end, we calculated the normalized

cross-correlation function (mean corrected) between them

and obtained its maximum (henceforth termed similarity

index). The similarity index is, therefore, sensitive to the

latencies of EEP excursions but not to overall EEP

amplitude. Again this analysis used the interval of

20–170 ms after stimulation onset. The template EEP was

calculated by averaging the EEPs that could be modified by

the dopamine D1-receptor antagonist SCH23390.

For most recording sites in auditory cortex, we com-

puted an auditory-evoked potential (AEP) by averaging all

LFP traces of a recording site relative to the onset of the

clicks (C50 trials).

Pharmacological modifications of local field potentials

in auditory cortex

To assess effects of the D1-receptor antagonist SCH23390

on EEPs in auditory cortex, we calculated correlation

coefficients (Pearson) between the preinjection EEP (i.e.,

the one measured before injection) and the EEPs measured

at several instances in time after injection. For this analysis,

again the time window of 20–170 ms after stimulation was

used. Subsequently, we considered all correlation coeffi-

cients \0.6 to indicate a significantly modified EEP. This

value was obtained from the control experiments in which

EEPs were repeatedly measured after injection of physio-

logical saline only or without any injection. In these control

experiments, correlation coefficients were found to be

rarely below 0.6 (less than 5 % of the cases).

In addition, we assessed whether the D1-receptor antag-

onist affected the ongoing LFP fluctuations in auditory cor-

tex. To this end, we calculated the median of the amplitude

spectrum within the delta/theta, alpha, beta, and gamma LFP

frequency bands (\6.7, 6.7–13.4, 20.1–26.9, 33.6–73.9 Hz)

from the fifty 152-ms periods before each of the electrical

stimulations. Subsequently, we compared the median of the

amplitude spectrum within each frequency band before

injection with each of those measured at different times after

injection using two-sided Wilcoxon rank-sum tests

(p \ 0.05, Bonferroni corrected for multiple comparisons).

Current source density analysis and determination

of cortical layers

In some experiments, we determined the laminar profile of

evoked potentials in auditory cortex (EEPs and AEPs) by

moving individual electrodes in steps of 100–300 lm along

tracks that were oriented almost perpendicular to the cor-

tical surface and computed the corresponding current

source density (CSD) profiles (Nicholson and Freeman

1975; Mitzdorf 1985; Steinschneider et al. 1992). The

laminar profile of evoked potentials was spatially smoothed

with an acausal, digital third-order low-pass Butterworth

filter (spatial cutoff frequency of 0.3 mm-1). Subsequently,

the second-order spatial derivative of the smoothed evoked

potentials was calculated using a 5-point Hamming win-

dow as described elsewhere (Happel et al. 2010). To retain

the upper and lower boundary sites, we used a linear

extrapolation method, which created two additional points

on each side outside of the measured range such that the

second derivative of the field potential assumed no addi-

tional decay (Happel et al. 2010).

For auditory stimulation, the laminar CSD profiles of

AEPs and the corresponding cascade of synaptic activa-

tions are well described (Steinschneider et al. 1992, 1998;

Happel et al. 2010; Kajikawa and Schroeder 2011). Since

current sources are thought to mainly reflect passive

compensatory currents, we considered only the sinks for

the interpretation, which for sensory stimulation were

shown to indicate active depolarizing events (Mitzdorf

1985). Sinks were defined as negative components (blue

colors in Fig. 5a, b) of the CSD profile if (1) they occurred

in at least two contiguous CSD traces and if (2) they were

[2 standard deviations below the mean baseline for at least

5 ms, computed from a 152-ms period before stimulus

onset. Depending on the spatial sampling density and the

extent of individual sinks usually one or two CSD traces

above and below the CSD trace with the maximal ampli-

tude of the sink were included for averaging. For each of

these averaged CSD traces of an individual sink, we

obtained the onset latency and the mean integral (i.e., the

charge density). The onset latency of a sink was defined as

the first point at which the sink was[2 standard deviations

(for C5 ms duration) below the baseline (see Kaur et al.

2005 or Happel et al. 2010).

Cortical layers of recording sites were determined by (1)

the depth relative to the cortical surface, (2) the presence

and absence of (auditory-evoked) firing, (3) the shape of

AEPs, and (4), if available, current sinks of auditory-

evoked CSD profiles.

Analyses of single and multiunit activity

For each single and multiunit record in auditory cortex, we

computed a post-stimulus time histogram (PSTH) with a

bin size of 10 ms relative to the onset of electrical stimu-

lation (C50 trials). Because electrical stimulation generated

an artifact that lasted up to 20 ms (as for LFP), this period

was excluded. To improve comparison between unit

activities recorded at different sites, we calculated relative

firing rates by dividing the PSTH through the average firing

rate during the 100-ms period before the onset of

stimulation.

Brain Struct Funct

123

To identify units that responded to deep brain stimula-

tion, we compared the firing rate in the 10-ms bin imme-

diately before stimulation with the firing rate in each of the

thirty 10-ms bins post-stimulus using Wilcoxon signed-

rank tests (two-sided). Firing of a site was considered to be

significantly modified if at least two consecutive of the

thirty bins were different from the bin immediately before

stimulation at p \ 0.0415, corresponding to an overall

significance level of 0.05 (0.0415 9 0.0415 9 29). Sig-

nificant decreases were considered to reflect inhibitory

responses; significant increases were considered to reflect

excitatory responses. For each PSTH, we obtained the first

bin with a significant difference (first-spike latency), the

number of significant bins (response duration), and the

firing rate and latency of the bin with the largest difference.

To assess the spectral sensitivity of a multiunit record,

we analyzed its responses to the tones presented at 40

different frequencies and obtained the best frequency, and

first-spike latency as described elsewhere (see Brosch et al.

1999).

Pharmacological modifications of multiunit activity

To uncover effects of the D1-receptor antagonist on mul-

tiunit activity in auditory cortex, we compared the PSTH

that was measured before drug injection with the PSTHs

that were measured at different times after injection. To

identify effects on electrically evoked multiunit responses,

we tested each PSTH for the presence or absence of such

responses. A multiunit was then considered to be affected

by the antagonist if an electrical response temporarily

disappeared after drug injection or if an electrical response

temporarily emerged at a previously unresponsive site. To

identify effects on the spontaneous firing, we analyzed

whether the spontaneous firing changed after injection of

the D1-receptor antagonist by means of Wilcoxon rank-

sum tests (two-sided, p \ 0.05, Bonferroni corrected for

multiple comparisons). The spontaneous firing was

obtained from the 100-ms period before electrical

stimulation.

Results

In total, we tested at 567 deep brain sites in three monkeys

(238, 157 and 172 in monkeys W, I, and E, respectively)

how electrical stimulation affected neuronal activity at

2,607 recording sites (1,202, 491, and 914 in monkeys W,

I, and E, respectively) in the primary auditory cortex, in the

secondary somatosensory cortex and in the superior tem-

poral polysensory cortex. The data were from 114 experi-

mental sessions in which we accessed the dopaminergic

ventral midbrain and surrounding deep brain structures by

moving stimulation electrodes along approximately dor-

soventrally oriented paths and in which we recorded LFPs

and neuronal discharges preferentially from the granular

layers of primary auditory cortex.

Influence of the dopaminergic ventral midbrain on local

field potentials in sensory cortex

The upper black line in Fig. 2a shows that brief monopolar

electrical stimulation of the dopaminergic ventral midbrain

evoked a polyphasic potential in auditory cortex that con-

sisted of several separable positive and negative deflec-

tions. In this experiment, electrode placement was based on

stereotactic coordinates, anatomical MRI measurements,

and a physiological landmark (i.e., the red nucleus). We

obtained additional evidence that the electrical stimulation

excited the dopaminergic ventral midbrain (i.e., the ventral

tegmental area or the substantia nigra) by finding that the

EEP could be modified by the D1-receptor antagonist

SCH23390. This is illustrated in Fig. 2a, which shows

color-coded EEPs that were repeatedly measured after

injection of the antagonist. We noted that from about

15–40 min after injection, a differently shaped EEP was

observed. After even longer periods EEPs were again

similar to that obtained before injection; thus the original

EEP was recovered. Note that at all times after injection an

EEP of some shape could be elicited, suggesting that the

dopamine antagonist only modulated neural transmission

from the dopaminergic ventral midbrain to auditory cortex,

but did not completely block it.

To quantitatively assess how long the dopamine antag-

onist affected the EEP, we calculated correlation coeffi-

cients (Pearson) to measure the relationship between the

EEP measured before injection with those measured after

injection (Fig. 2a, right panel). This analysis revealed that

correlation coefficients remained at high levels until

10 min after injection, and then markedly decreased and, at

some time, even attained negative values; the latter indi-

cates substantial anticorrelation and, thus, partially oppo-

site phase relationships between the EEPs obtained before

and after injection. After being most negative 30 min after

injection, the correlation coefficient increased again and

then remained at high levels from about 45 min after

injection.

To determine which changes of correlation coefficients

indicated a statistically significant effect of the antagonist

on the EEP, we assessed the variance of such EEPs when

they were repeatedly measured without any pharmacolog-

ical intervention. As shown for another stimulation/

recording pair (Fig. 2b), we found that in this condition the

EEP was highly reproducible over a period of 90 min. This

was also seen for another 18 sites that were estimated to be

located in the dopaminergic ventral midbrain and that had

Brain Struct Funct

123

similar EEPs. The high stability of EEPs was reflected in

generally high correlation coefficients that were only rarely

(\5 % of all cases) below 0.6. Thus, we considered

decreases of correlation coefficients below this critical

value to indicate a statistically significant change of an

EEP. In the example shown in Fig. 2a, we therefore con-

cluded that the dopamine antagonist affected the EEP from

15 to 40 min after injection.

Population analysis of electrically evoked potentials

in auditory cortex

At 44 of the deep brain sites (including the one shown in

Fig. 2a), we determined whether dopamine was involved in

the transmission of neuronal signals to the auditory cortex

by injecting the D1-receptor antagonist SCH23390. For 28

(63.6 %) of these sites, we found that SCH23390 signifi-

cantly modified the EEP recorded in auditory cortex (i.e., at

least one of the correlation coefficients computed from the

preinjection EEP and the postinjection EEP was below the

critical value of 0.6), with no large difference between

injections made intramuscularly (21/31) and into auditory

cortex (7/13). This was found at 135 of 212 sites (63.7 %)

in auditory cortex. Modifications occurred between 5 and

85 min after injection and were most prevalent 25–40 min

after injection (Fig. 2c). Notably, EEP changes were

exclusively observed when deep brain sites were located in

the dopaminergic ventral midbrain (28/41). By contrast, no

such changes were observed for EEPs obtained from sites

[4 mm dorsally from the dopaminergic ventral midbrain

(0/3), which also had very different EEP shapes (e.g., see

Fig. 3a). This suggests that the shape of the EEP elicited

from the dopaminergic ventral midbrain is distinct from the

shapes of EEPs that were elicited from locations outside

this deep brain structure. This observation was, therefore,

used as an additional means to determine whether a deep

brain site that was not pharmacologically tested was inside

or outside the dopaminergic ventral midbrain, namely to

electrically stimulate this site and determine whether the

resulting EEP was similar to the EEP that is specific for the

dopaminergic ventral midbrain. To assess this similarity,

we calculated the normalized cross-correlation function

between an EEP and a template EEP that is specific for the

dopaminergic ventral midbrain. The template EEP is the

average of the 28 EEPs that were modified by the D1-

receptor antagonist SCH23390, and consists of several

positive and negative deflections with amplitudes of -175

to 130 lA within a time range of about 400 ms (Fig. 2d).

Because its most prominent and reproducible deflections

were in the time window of 20–170 ms after stimulation

only this time window was used for the calculation of the

cross-correlation function. For further purposes, we

obtained the maximum from the cross-correlation function

and considered all EEPs to match the template EEP if this

similarity index exceeded 0.8.

The observed changes of EEPs after injection of

SCH23390 are compatible with the interpretation that the

EEP results from a superposition of postsynaptic potentials

(Jervis et al. 1983) that is modified by the D1-receptors.

The alternative view of EEP generation is that the volley of

action potentials arriving after electrical stimulation resets

the phase of spontaneous cortical oscillations so that

activity at particular frequencies becomes phase locked

(Sayers et al. 1974). This view is not compatible with the

effects of SCH23390 on spontaneous LFP oscillations.

Specifically, we tested whether LFP amplitudes in the

delta/theta, alpha, beta, and gamma frequency bands (\6.7,

6.7–13.4, 20.1–26.9, 33.6–73.9 Hz) changed after injection

of SCH23390. The spontaneous LFP fluctuations were

obtained from the 152-ms periods before the onsets of

electrical stimulation trains. Although there were signifi-

cant (Wilcoxon rank-sum tests, two-sided, p \ 0.05)

changes found for individual deep brain sites the incidence

of these changes was not above the chance level obtained

in our control experiments.

Characterization of electrically evoked potentials

elicited from dopaminergic ventral midbrain

and adjacent deep brain sites

Figure 3a shows examples of the different EEP shapes

measured in granular layer of auditory cortex that were

elicited from different deep brain sites by moving a stim-

ulation electrode along an approximately ventromedially

oriented path into the ventral tegmental area. In this track,

the red nucleus could be unequivocally identified because

electrical stimulation evoked body movements that are

specific for this nucleus (Lovell et al. 2014). After failing to

evoke these body movements, we concluded that the

electrode had reached the ventral tegmental area. We noted

that from every site along the entire track an EEP could be

elicited. Their shapes and amplitudes, however, varied

considerably. For the current study, it is of interest that the

EEPs elicited within the range from 23.5 to 29.75 mm

below cortical surface were all highly similar to each other

and had similar amplitudes. This range corresponds to the

location of the ventral tegmental area. That EEPs within

this range were elicited from the ventral tegmental area

was further indicated by finding that all EEPs had high

similarity indices ([0.8), i.e., they were highly similar to

the template EEP, which is specific for the dopaminergic

ventral midbrain (see above). By contrast, EEPs elicited

from outside this range (e.g., the red nucleus) had small

similarity indices (\0.5) and larger amplitudes.

From 381 of the 523 deep brain sites (72.9 %) that were

not pharmacologically tested, we elicited EEPs at 1,748 of

Brain Struct Funct

123

2,395 (73.0 %) recording sites in auditory cortex with a

significant similarity index ([0.8), i.e., the EEPs that could

be elicited from them matched very well the EEP of the

sites where the D1-receptor antagonist could modify the

EEP. Sites with a significant similarity index were prefer-

entially located within a range of *9 mm centered around

27 mm below cortical surface. Within this range[50 % of

the EEPs had a similarity index [0.8 (Fig. 3c), and the

median similarity index was [0.8 (Fig. 3b). These spatial

measures are in good correspondence with the location of

the ventral tegmental area and the substantia nigra based on

physiological landmarks and published brain atlases (Szabo

and Cowan 1984; Martin and Bowden 1996). Within the

range of high similarity indices, EEP amplitudes were

relatively constant (expressed as median RMS values in the

time range of 20–170 ms after stimulation onset, Fig. 3d).

Fig. 3 EEPs elicited from the dopaminergic ventral midbrain differ

from those elicited from other deep brain sites. a Examples of the

EEPs that were elicited from different deep brain sites by moving a

stimulation electrode along a track that was oriented at an angle of

28� from the horizontal plane. Depth was measured relative to cortical

surface. In the heat plot, each row shows the color-coded amplitudes

of the EEP that was elicited from the depth given on the ordinate.

Amplitudes above and below 150 lV were clipped. The four black

lines show the EEPs from selected deep brain sites, such as the

dopaminergic ventral midbrain, i.e., the ventral tegmental area (VTA)

and the substantia nigra (SN), the red nucleus (RN), the midbrain

reticular formation (MRF) and unidentified sites, separated by

horizontal dotted lines. Note that EEPs elicited from within the

dopaminergic ventral midbrain (VTA/SN) were highly similar to each

other but different from those elicited from outside the VTA/SN. For

illustrative purposes, the depth profiles of two adjacent electrode

tracks were pooled and smoothed with a spatial constant of 400 lm.

The electrode track corresponds to a track similar to the one shown in

Fig. 1b. b–c Variation of median similarity index of EEPs (b) and the

percentage of EEPs that were similar to the template EEP (c) with

depths of the stimulation sites for all monkeys (main panel) and for

individual monkeys (insets). The similarity index is defined as the

maximum of the cross-correlation function between an individual

EEP and the template EEP (see Fig. 2d). An EEP was considered to

be similar to the template EEP if the similarity index was C0.8.

d Variation of the median amplitudes (root mean square) of EEPs

with depth of the stimulation site. Root mean squares were calculated

from the time range of 20–170 ms after stimulation onset. Insets

indicate results of the three monkeys individually

Brain Struct Funct

123

Electrically evoked potentials in different layers

of auditory cortex and in the somatosensory cortex

and the superior temporal polysensory cortex

For 17 sites in the dopaminergic ventral midbrain, we

tested how electrical stimulation affected LFPs in auditory

cortical layers other than the granular one and in other

sensory cortices. To this end, we placed stimulation

electrodes at fixed positions in the dopaminergic ventral

midbrain and measured EEPs at different depths of the

auditory cortex, the somatosensory cortex and the superior

temporal polysensory cortex by moving recording elec-

trodes along 63 trajectories that were oriented approxi-

mately perpendicular to the surface of auditory cortex.

Examples of the EEPs that were measured along such an

electrode track are given in Fig. 4a. We noticed that, within

Fig. 4 Highly similar EEPs in different layers of the auditory cortex

and in the somatosensory cortex and the superior temporal polysen-

sory cortex. a Examples of the EEPs that were elicited at different

sites of the auditory cortex, the somatosensory cortex and the superior

temporal polysensory cortex by stimulating a location in the

dopaminergic ventral midbrain. The cortical recordings were obtained

by moving an electrode along a trajectory that was oriented slightly

off the dorsoventral plane (10�). Depth is given relative to the

granular layer of auditory cortex. In the heat plot, each row shows the

color-coded amplitudes of the EEP that was measured at the depth

given on the ordinate. Amplitudes above and below 180 lV were

clipped. The five black lines show the EEPs at selected depths.

Approximate locations of the secondary somatosensory cortex (S2),

the lateral sulcus (ls), the supragranular (SG), granular (G), and

infragranular (IG) layer of auditory cortex, the white matter, and the

superior temporal polysensory cortex (STP) are indicated right of the

heat plot. b–c Variation of median similarity index of EEPs (b) and

the percentage of EEPs that were similar to the template EEP (c) with

depths of the recording sites for all monkeys and for individual

monkeys (insets). The similarity index is defined as the maximum of

the cross-correlation function between an individual EEP and the

template EEP (see Fig. 2d). An EEP was considered to be similar to

the template EEP if the similarity index was C0.8. d Variation of

median amplitudes (root mean square) of EEPs with depth of

recording site. Root mean squares were calculated from the time

range of 20–170 ms after stimulation onset. Insets show the two

monkeys individually. Red dots indicate statistically significant

measures (see ‘‘Methods’’)

Brain Struct Funct

123

the investigated time window of 20–170 ms post-stimulus,

the major excursions of the EEPs were very similar at all

depths and thus in all cortical layers of auditory cortex. At

later times, EEP excursions were more variable. In addi-

tion, we noticed that major excursions of the EEPs were

very similar in the auditory, somatosensory, and superior

temporal polysensory cortex. These observations were

quantitatively confirmed by finding that all EEPs in this

track had high similarity indices ([0.8).

For the remaining 16 sites in the dopaminergic ventral

midbrain, we confirmed that EEPs were highly similar at

different depths of the auditory and the somatosensory

cortex as well as of the superior temporal polysensory

cortex. Neither the similarity index of the EEPs (Fig. 4b)

nor the percentage of EEPs with high similarity to the

template EEP (Fig. 4c) varied significantly with cortical

depth [bootstrap, p [ 0.05. For the bootstrap procedure, we

iteratively (100,000 times) obtained from all 17 stimulation

sites the variation of the random depth profile for the

similarity index (or the number of template matched EEPs)

by randomly shuffling with replacement the cortical depth/

similarity index pairs of a stimulation site]. Despite con-

siderable variation across individual tracks, the amplitudes

of the EEPs were generally quite similar at different sites of

the auditory and the somatosensory cortex as well as of the

superior temporal polysensory cortex. Although our boot-

strap analysis was not able to reject such a dependence

(p \ 0.05), a significant difference was present at a single

cortical depth only, which according to additional analyses

(see below), corresponded to the supragranular layer of

auditory cortex (Fig. 4d).

Laminar current source density profiles of electrically

evoked potentials in auditory cortex

Although EEP excursions appeared at similar latencies at

different depths of auditory cortex, when the interval from

20 to 170 ms after stimulation was considered, amplitudes

of individual excursions were quite variable. This is

exemplified in Fig. 5a in which EEPs were sampled with a

higher spatial density than in Fig. 4a. The variability of

EEP amplitudes is best visible in the color-coded repre-

sentation of the EEPs, which demonstrates that some of the

later EEP peaks after 170 ms, even changed polarity with

cortical depth, particularly the one around 220 ms. To

better assess the layer-specific activation of auditory cor-

tex, we, therefore, performed a CSD analysis on the EEPs

that were measured at different cortical depths. To tune up

our method and to have a better estimate of the cortical

layer from which LFPs were recorded, we also performed a

CSD analysis on click-evoked potentials.

The auditory-evoked CSD profiles were similar to those

obtained with click and tone stimulation in previous studies

on the auditory cortex of monkeys (Steinschneider et al.

1992, 1998; Kajikawa and Schroeder 2011) and rodents

(Happel et al. 2010, 2014). Earliest click-evoked current

sink activation was found in the ventral most recordings

from auditory cortex, i.e., in deeper layers, with a median

onset latency of 28.0 ms (interquartile range,

IQR = 1.7 ms) after stimulation. This was followed by a

sink (60.0 ms, IQR = 0.2 ms) in the middle portions of

auditory cortex. Subsequent sink activations were found

again in deeper layers around 126 ms (IQR = 6.6 ms) and

in upper parts (200 ms, IQR = 2.1 ms). The last sink in

upper parts was usually the most prominent one, i.e., the

one with the greatest charge density. This spatiotemporal

cascade of sinks was found in 34 of 35 CSD profiles which

we obtained during 8 experimental sessions in three mon-

keys (Fig. 5c, dark gray bars). Based on the similarity of

our findings with those obtained in the cited earlier studies,

we assigned recordings with early and late sinks in the

ventral most parts of auditory cortex to infragranular lay-

ers, recordings with a sink in the middle parts of auditory

cortex to granular layers, and recordings with a sink in the

dorsal most parts of auditory cortex to supragranular layers.

For the interpretation of laminar EEP profiles, we used

the same reasoning that has been established for laminar

AEP profiles, namely that only current sinks indicate syn-

aptic activations and should therefore be described (Mul-

ler-Preuss and Mitzdorf 1984; Mitzdorf 1985;

Steinschneider et al. 1992; Happel et al. 2010). The rep-

resentative example in Fig. 5a shows that electrical stim-

ulation in the dopaminergic ventral midbrain evoked a

highly specific cascade of current sinks across cortical

layers. The same cascade of four sinks was present in

almost all CSD profiles (59 of 63), which we obtained

during 17 experimental sessions in two monkeys. The

cascade of current sinks usually started with an early in-

fragranular sink at 42 ms (IQR = 3.0 ms), which was

followed by a granular sink at 95 ms (IQR = 1.7 ms), by

another late infragranular sink at 132.5 ms

(IQR = 5.7 ms), and a supragranular sink at 235.5 ms

(IQR = 12.3 ms) (Fig. 5c, light gray bars). Compared to

the timing relationships of the sinks, there was more var-

iation of the strengths of the different sinks among the 63

electrically evoked CSD profiles (Fig. 5d). The largest

sinks (expressed as the charge density, i.e., the integral of

the current flow during the presence of a sink) were the late

infragranular and the supragranular sinks, and the weakest

sink was the early infragranular sink (Fig. 5d, light gray

bars).

Interestingly, the same spatiotemporal cascade of sinks

was observed when auditory instead of electrical stimuli

were tested, that is, an early infragranular sink was fol-

lowed by a granular sink, a late infragranular sink and a

supragranular sink (Fig. 5b). We noticed that for the

Brain Struct Funct

123

stimulation parameters used in the present study auditory

sinks were consistently earlier and had larger amplitudes

than electrical sinks. It is not clear whether these rela-

tionships change if different auditory or electrical stimuli

would be tested. The similarities in the temporal order of

the intracolumnar signal flow was unlike the strengths of

synaptic activations evoked by auditory and electrical

stimulation, which was strongest in supragranular layers,

followed by that of granular layers and weakest for the

early infragranular sink (Fig. 5d, dark gray bars).

Influence of the dopaminergic ventral midbrain

on neuronal firing in auditory cortex

Electrical stimulation of the dopaminergic ventral midbrain

could also affect neuronal firing in auditory cortex. This

was found when we analyzed 235 sites in this deep brain

structure. They were from the 409 sites from which an EEP

could be elicited that was highly similar to the template

EEP (similarity index [0.8; the remaining 174 sites were

excluded for technical reasons). For each of the 235 deep

Fig. 5 Electrically and acoustically evoked potentials in different

layers of auditory cortex and corresponding current source density

profiles. a Representations of EEPs recorded at different depths of

auditory cortex by lines (left panel) and by a heat plot (middle panel)

as well as corresponding second spatial derivatives of EEPs (CSD

profile; right panel). Note the layer-specific activation sequence of

current sinks (blue colors), starting in upper parts of the infragranular

layer (IGe), followed by those in granular layers (G), deeper parts of

the infragranular layers (IGl) and in supragranular layers (SG).

b Depth profile of AEPs and corresponding CSD profile. Note that the

activation sequence of current sinks was similar to that for the EEP

shown in a and that the range of recording depths is greater than in

a. c Median and interquartile range (IQR) onset latencies of the four

sinks obtained from the CSD profiles for electrical (ES; light gray

bars) and acoustical (AS; dark gray bars) stimulation. d Median and

IQR charge densities of the four sinks obtained from the CSD profiles

Brain Struct Funct

123

brain sites, we analyzed the effects of stimulation on up to

seven sites in auditory cortex. The number of cortical sites

was variable because only multiunit recordings were con-

sidered at which the neurons responded to tones or clicks

(755) and/or responded to electrical stimulation (140, see

below), such that a total of 769 multiunit recordings in

auditory cortex entered our analyses.

Figure 6 shows two example sites in auditory cortex

where the multiunit activity was briefly increased (panel A)

or decreased (panel B) for tens of milliseconds shortly after

electrical stimulation of the dopaminergic ventral mid-

brain. Such phasic responses were seen at 140 of the 769

sites in auditory cortex (18.2 %) and could be elicited from

89 of 235 (38 %) sites in the dopaminergic ventral

Fig. 6 Single and multiunit responses in auditory cortex evoked by

electrical stimulation of the dopaminergic ventral midbrain. a Repre-

sentative multiunit recording with an excitatory response. In the raster

display, each dot in a horizontal raster line indicates the occurrence of

a spike relative to onset of electrical stimulation. Different raster lines

refer to the 50 repetitions of electrical stimulation. The lower panel

shows the average firing rate (PSTH) relative to the onset of electrical

stimulation. The gray-shaded area indicates the time interval during

which the firing rate was significantly different from the firing rate in

the 10-ms bin immediately before stimulation onset (Wilcoxon

signed-rank test, two-sided). b Representative multiunit recording

with an inhibitory response. c Gray-coded plot of the 74 multiunits

with excitatory electrical responses. Each row shows the time course

of the relative firing rate of an individual multiunit encoded on a

grayscale. d Gray-coded plot of the 66 multiunits with inhibitory

electrical responses. e–f Population PSTH of the 42 single units with

excitatory responses (e) and the 15 single units with inhibitory

(f) responses

Brain Struct Funct

123

midbrain. Unlike all EEPs, electrical multiunit responses

were (with 1 exception) always monophasic. About half of

the responses (n = 74, 52.9 %) were excitatory. The

remaining 66 responses (47.1 %) were inhibitory ones. The

characteristics of all electrical multiunit responses are

shown in panels C and D of Fig. 6, respectively. Generally,

excitatory responses occurred significantly earlier than

inhibitory responses after stimulation (Wilcoxon rank-sum

tests, two-tailed, p \ 0.0034). Median first-spike latencies

of excitatory responses were 35.0 ms (IQR = 4.3 ms),

those of inhibitory responses 50.0 ms (IQR = 8.0 ms). The

median time of maximal excitation was 75 ms

(IQR = 7.6 ms) and that of maximal inhibition was

130 ms (IQR = 14.9 ms). At these times, excitatory

responses were increased to 160 % (IQR = 10 %) above

the spontaneous firing; inhibitory responses were reduced

to 50 % (IQR = 3 %) of the spontaneous firing. Median

duration of both types of electrical responses were equal

(60.0 ms; IQR = 4.3 ms and IQR = 5.7 ms, respectively,

Wilcoxon p [ 0.05).

The sites with electrical responses had auditory response

properties that did not differ from those of cortical sites that

did not respond to electrical stimulation. Auditory response

properties analyzed were best frequency and response

latency.

We also obtained evidence that electrical stimulation of

the dopaminergic ventral midbrain evoked excitatory and

inhibitory responses in single units in auditory cortex. This

was found when we analyzed 57 single units which were

isolated from the 140 sites where electrically evoked

responses were observed in the multiunit activity. Inspec-

tion of the PSTHs triggered on the onset of the electrical

stimulus suggested weak electrically evoked responses in

several single units with none being statistically significant.

To quantitatively describe electrically evoked responses in

single units nevertheless, we generated two population

PSTHs from the firing of the 57 single units. The first

population consisted of the 42 single units that were

recorded from sites where an excitatory electrically evoked

multiunit response was detected (Fig. 6e). A Wilcoxon test

(two-sided, p \ 0.05) revealed an excitatory electrical

response lasting from 20 to 140 ms after stimulus onset.

The second population consisted of the 15 single units

where an inhibitory electrically evoked multiunit response

cFig. 7 Temporary modification of electrically evoked multiunit

responses in auditory cortex by blocking the dopamine D1-receptor.

a Temporary suppression of the excitatory electrical response. Each

row in the upper panel shows the time course of the firing rate in

response to electrical stimulation (on a gray scale) before and at

several points in time after injection of the D1-receptor antagonist

SCH23390. Note the overall decrease of firing and the loss of the

excitatory electrical response from 25 to 55 min after injection. Stars

indicate significant changes of the evoked firing. The lower panel

shows the average firing rate (PSTH) of this multiunit relative to the

onset of electrical stimulation before injection (light blue curve),

25–55 min after injection (red curve), and from 65 min after injection

(dark blue). b Temporary emergence of an inhibitory electrical

response at *62 ms (red dot). Note the overall increase of firing and

the emergence of the inhibitory electrical response from 5 to 55 min

after injection. Also, note that the initial peak at *37 ms, superim-

posed on the inhibitory response, does not exceed the baseline firing

rate and is, therefore, not considered an excitatory response in this

multiunit recording

Brain Struct Funct

123

was detected (Fig. 6f). A Wilcoxon test revealed an

inhibitory electrical response lasting from 70 to 200 ms

after stimulus onset.

Electrically evoked multiunit responses are affected

by the D1-receptor antagonist SCH23390

Like the EEPs, we found that multiunit responses to elec-

trical stimulation of the dopaminergic ventral midbrain

were under control of D1-receptors. Blocking them with

the antagonist SCH23390 could attenuate or amplify such

electrical responses in auditory cortex. These effects were

seen both with intramuscular and intracortical injections.

Figure 7a shows a site with an excitatory response to

electrical stimulation of the dopaminergic ventral midbrain

(light blue curve). This response became smaller or even

disappeared shortly after injection of the D1-receptor

antagonist (red curve). The effect of the antagonist lasted

for 30 min after which the excitatory response was

recovered (dark blue curve). Note that the changes of the

electrical response were accompanied by a temporary

reduction of the spontaneous firing of the multiunit.

Figure 7b shows an example in which the D1-receptor

antagonist had the opposite effect on the neural transmis-

sion from the dopaminergic ventral midbrain to auditory

cortex. The neurons at this site usually did not respond to

electrical stimulation (light blue curve). Within 5 min after

injection, however, electrical stimulation was able to evoke

an inhibitory response (according to our criterion) and this

could be repeated during a period of 50 min (red curve),

after which no more inhibitory responses could be evoked

(dark blue curve). Note that also in this example the

antagonist changed the spontaneous firing of the multiunit,

however, it increased it. Presumably only this increase in

spontaneous firing made it possible that the inhibitory

action from the dopaminergic ventral midbrain was able to

affect the (suprathreshold) firing.

The two types of effects of the antagonist SCH23390 on

electrically evoked multiunit responses were seen at 39

sites in auditory cortex, including the two examples shown

in Fig. 7, with no obvious differences between intramus-

cular and intracortical injections. This was found when we

analyzed multiunit activity at 184 sites primarily in the

granular layer of auditory cortex while electrical stimula-

tion was applied at 47 sites in the dopaminergic ventral

midbrain. Twenty sites in auditory cortex at which elec-

trical responses were evoked under normal conditions lost

their responses after injection of the antagonist. This was

observed both for sites with excitatory responses (7) and

for sites with inhibitory responses (13). The antagonist had

no effect on six sites in auditory cortex at which electrical

responses were evoked under normal conditions. At

another 19 sites in auditory cortex, we observed that

electrical responses could temporarily be evoked after

injection of the antagonist. Most of them were inhibitory

responses (13), a smaller fraction were excitatory responses

(6). No such emergence of responses was seen in 22

multiunits in auditory cortex at which we tested physio-

logical saline instead of SCH23390. The time window

during which the D1-receptor antagonist SCH23390 was

effective was similar to that for EEPs, i.e., it was in the

range of 5–60 min after drug injection. This is consistent

with previous observation in primate prefrontal cortex

(Sawaguchi and Goldman-Rakic 1994). At many sites, the

change of an electrical response was accompanied by a

change of the spontaneous firing, mostly an increase. These

changes could also occur at sites that did not respond to

electrical stimulation in any condition.

Discussion

The current study on awake macaque monkeys shows for

the first time that electrical stimulation of the dopaminergic

ventral midbrain resulted in short-latency phasic activation

of auditory cortex. Activations were reflected in local field

potentials and spikes recorded in all cortical layers of

auditory cortex and were also seen in the secondary

somatosensory cortex and the superior temporal polysen-

sory cortex. This shows that these cortices are under con-

trol of dopaminergic cell structures in the ventral midbrain.

In the following, we discuss that these activations are most

likely generated by non-dopaminergic transmitters, that

dopamine modulates this transmission, and that they pro-

vide one source for the representation of non-auditory

aspects of auditory tasks in the auditory cortex.

Spatial spread of electrical stimulation

In this study, placement of stimulation electrodes was

guided by a combination of stereotactic and physiological

criteria. These included observing firing that is character-

istic of dopaminergic neurons (e.g., Grace and Bunney

1983, 1984a, b) and finding neighboring brain nuclei with

characteristic physiological properties, such as the red

nucleus (Lovell et al. 2014). At selected locations, we

verified that electrical stimulation excited dopaminergic

neurons by demonstrating that the EEP in auditory cortex

was modified by the D1-receptor antagonist SCH23390.

These tests also revealed that stimulating different loca-

tions of the dopaminergic ventral midbrain elicited EEPs

that were highly similar to one another. Stimulation of sites

outside the dopaminergic ventral midbrain either elicited

no EEP or an EEP of different shape. These could be

elicited from auditory structures like medial geniculate

body or from structures classically not considered part of

Brain Struct Funct

123

the auditory system. For example, Fig. 3a shows an EEP of

a different shape that was elicited from the red nucleus and

which we have considered to be indicative of a functional

connection to the auditory cortex (Lovell et al. 2014).

Thus, we concluded that the shape of the EEP can be used

as an additional means for correct placement of an elec-

trode in the dopaminergic ventral midbrain.

When we applied our criteria, we found EEPs with a

shape that was specific for the dopaminergic ventral mid-

brain within a range extending about 9 mm in dorsolateral–

medioventral direction. This matched very well with the

published coordinates of the ventral tegmental area and

substantia nigra. It is reasonable to assume that the stim-

ulation parameters standardly used in the present study

(100 lA, monopolar, anodal first) excited neurons mainly

within a radius of less than 1 mm (Ranck 1975; Histed

et al. 2009). This estimation is supported by finding clear

boundaries between different deep brain structures. For

example, there was an abrupt change of the EEP shapes at

23.25 mm, when the stimulation electrode was moved

from the red nucleus into the ventral tegmental area (see

Fig. 3a). Thus, our stimulations were quite focal and in

most cases, confined to one of the two dopaminergic

nuclei. Within the two dopaminergic nuclei, the electrical

stimulus likely excited dopaminergic, glutamatergic, and

GABAergic neurons (and possibly also fibers of several

brain stem nuclei that pass through the ventral tegmental

area), with no obvious differences of EEPs between ventral

tegmental area and substantia nigra. This was unexpected

for us given that the two dopaminergic nuclei are generally

considered to be part of distinct projection systems, namely

the mesocortical and the nigrostriatal pathway (Oades and

Halliday 1987). It is possible that this distinction is not as

strict for auditory cortex because a recent study on rodents

found that auditory cortex receives more projections from

the substantia nigra than from the ventral tegmental area

(Budinger et al. 2008). Because of the short latencies of the

electrically evoked activations in auditory cortex and the

conduction velocity of the mesocortical pathway (Deniau

et al. 1980; Thierry et al. 1980), and because all cell types

participate in the mesocortical projection to the prefrontal

cortex (Thierry et al. 1973; Steffensen et al. 1998; Hur and

Zaborszky 2005) we concluded that the earliest electrically

evoked activations of auditory cortex are likely due to the

monosynaptic connections from the dopaminergic ventral

midbrain to auditory cortex. The effects found after

injection of the D1-receptor antagonist (Fig. 2) argue

against that electrically evoked responses arise from anti-

dromic activation of cortical fibers projecting to the

dopaminergic ventral midbrain or from activations of fibers

of passage. This is consistent with more extended testing

performed on the mesocortical system in rats (Lavin et al.

2005).

Comparison of electrically evoked responses

in different cortical areas

Our study is in good agreement with previous studies on

prefrontal cortex that electrical stimulation of the ventral

tegmental area and substantia nigra results in short-latency

and short-lasting neuronal activations of cortex and that

this is reflected in postsynaptic potentials (Bernardi et al.

1982; Ferron et al. 1984; Mercuri et al. 1985; Lavin et al.

2005), in extracellular field potentials (Lavin et al. 2005)

and in neuronal firing (Pirot et al. 1992; Au Young et al.

1999; Pistis et al. 2001; Lavin et al. 2005). In the following,

we argue that most of the differences of electrically evoked

responses in prefrontal and auditory cortex reside in

methodological and species differences and not in differ-

ences how the dopaminergic ventral midbrain affects

neuronal activity in different cortical regions.

Latencies of electrically evoked responses were longer

in sensory cortex than in prefrontal cortex. One reason for

this is that our measurements were made in monkeys with

larger brains and thus larger transmission times than those

of rodents. Another and possibly more important reason is

the effect of anesthesia, which reduces response latencies

and which was used in all studies of prefrontal cortex, but

not in the current study. For example, median minimum

latency of acoustic responses in primary auditory cortex

was 12 ms shorter in anesthetized animals than in awake

animals (Ter-Mikaelian et al. 2007). Therefore, the anes-

thesia-induced reduction of response latencies could

account for latency differences between prefrontal cortex

and sensory cortex. It should also be noted that the stimulus

artifact prohibited observation of latencies below 20 ms in

our study.

The choice of stimulation parameters may explain why

electrically evoked responses consisted of fewer waves in

prefrontal cortex than in sensory cortex. For instance, La-

vin and colleagues (Lavin et al. 2005) found that using a

short pulse train instead of a single pulse, evoked potentials

in prefrontal cortex with more than the maximally two

components that have typically been reported in the pre-

frontal studies.

Similar to prefrontal cortex, stimulation of the dopami-

nergic ventral midbrain resulted in phasic increases and

decreases of the neuronal firing in auditory cortex. These

responses lasted for similar periods of time, with excitatory

responses occurring earlier than inhibitory responses and

every site (with 1 exception) exhibiting either an excitatory

or an inhibitory response only. A difference, however, was

that we found equal numbers of the two types of neuronal

responses whereas in the prefrontal cortex, inhibitory

responses prevailed. One reason for this may be that dif-

ferent cortical cell types show different electrical respon-

ses. Tseng et al. (2006) showed that pyramidal cells

Brain Struct Funct

123

responded with an inhibition after activating the ventral

tegmental area with a small NMDA injection, whereas fast-

spiking interneurons increased firing after such an injec-

tion. Because our study was based on multiunit activity,

our data include both pyramidal cells and other cell types,

such as cortical inhibitory interneurons. By contrast most

of the studies on prefrontal cortex were likely biased

toward pyramidal cells because these cells fire large action

potentials that are more easily isolated in extracellular

single unit recordings.

Our study also agrees with previous work on the mes-

ocortical system that electrical responses in cortex are

under dopaminergic control. For instance, we found that

the D1-receptor antagonist SCH23390 modified the shape

of EEPs and the strength of evoked firing in auditory

cortex. This is consistent with the study of Lavin and

colleagues (2005) who found that after local destruction of

dopaminergic cells with 6-hydroxydopamine, stimulation

of the ventral tegmental area failed to evoke postsynaptic

potentials in the prefrontal cortex. This finding was con-

firmed for the electrically evoked depolarizations (but not

for the subsequent hyperpolarizations) that were observed

by optical imaging of neuronal activity with a voltage-

sensitive dye (Watanabe et al. 2009). The cited studies also

tested other transmitter systems and identified that the

initial fast excitatory response was mediated by glutamate.

Further tests suggested that this transmitter was probably

coreleased by the dopaminergic terminals in cortex and not

released from glutamatergic projection neurons in the

ventral tegmental area (Lavin et al. 2005). Lavin et al.

(2005) also found evidence that the following inhibitory

potential was mediated by GABA.

Because of the large similarities of the electrically

evoked responses in the auditory cortex, the somatosensory

cortex, the superior temporal polysensory cortex (this

study), and prefrontal cortex (e.g., Lavin et al. 2005; Wa-

tanabe et al. 2009), we consider that the electrically evoked

responses in sensory cortex are also generated by glutamate

that is coreleased from the dopaminergic terminals and by

subsequent GABA-mediated inhibition. This is consistent

with the conclusions of Seamans and Yang (2004) that the

postsynaptic effects of dopamine are too slow to generate

EEPs with a short-latency and a short duration. Neverthe-

less, one has to take into account that the density of

dopaminergic fibers falls off from rostral to caudal parts of

the cortex, thus, that the mesocortical pathway likely has a

stronger effect on the prefrontal cortex than on the sensory

cortex (Swanson 1982; Bjorklund and Lindvall 1984;

Lewis et al. 1987; Berger et al. 1991; Durstewitz et al.

1999).

In conclusion, electrical stimulation in the dopaminergic

ventral midbrain appears to similarly affect neuronal

activity in different regions of the auditory cortex, the

somatosensory cortex, the superior temporal polysensory

cortex, and the prefrontal cortex. Whether this also applies

to other cortical regions is brought into question by an

imaging study using voltage-sensitive dyes. It was reported

that electrical stimulation in the ventral tegmental area

evoked changes in optical signals only within an entire

field of prefrontal cortex but spared adjoining primary

motor cortex and more remote primary somatosensory

cortex (Watanabe et al. 2009).

Layer-specific activation of auditory cortex

by electrical stimulation

Our CSD analysis suggests that the volley of action

potential evoked by electrical stimulation of the dopami-

nergic ventral midbrain initiates a stereotypic layer-specific

activation sequence in auditory cortex. It is initially most

pronounced in the infragranular layer and subsequently

proceeds from the granular layer back to the infragranular

and to the supragranular layer. Because the initial part of

the EEP is most likely mediated by glutamate (see above),

our results suggest that the density of the mesocortical

glutamate (co)release sites is highest in the infragranular

layer of auditory cortex. Although to our knowledge this

question has not yet been addressed anatomically, it is

interesting that this corresponds to reports that the density

of dopaminergic terminals is highest in infragranular layers

(Campbell et al. 1987; Lewis et al. 1987). The higher

density of the dopaminergic terminals in the infragranular

layers entails an increased probability for the occurrence of

glutamate release sites.

To our surprise, the sequence of activations evoked by

electrical stimulation of the dopaminergic ventral midbrain

mirrored the sequence of activations in auditory cortex

evoked by auditory stimulation. The delayed progression of

the intracolumnar signal flow after electrical stimulation

might simply reflect differences in the strength of stimu-

lation. The similarities of the spatial cascades might be due

to the recruitment of the same transmitter systems after

auditory and electrical stimulation. They also suggest that

the later electrically evoked activations in auditory cortex

reflect a mixture of currents arising from intracolumnar and

extracolumnar inputs (Happel et al. 2010, 2014) and from

polysynaptic inputs from the dopaminergic ventral

midbrain.

Functional implications

The current electrical stimulation study indicates that

fluctuations of neuronal activity in the dopaminergic ven-

tral midbrain are rapidly conveyed to auditory cortex and

other parts of sensory cortex. Here they depolarize and

sometimes also hyperpolarize neurons in all layers for up to

Brain Struct Funct

123

a few hundred milliseconds. In a substantial fraction of

these neurons, the changes of membrane potentials are

suprathreshold and even affect their firing. Thus, the

mesocortical projection may contribute to the synchroni-

zation of neurons in the dopaminergic ventral midbrain

with auditory, secondary somatosensory, or superior tem-

poral polysensory cortex, as has been demonstrated for the

prefrontal cortex (Gao et al. 2007; Fujisawa and Buzsaki

2011).

On a functional level our findings suggest that informa-

tion carried by neurons in the dopaminergic ventral midbrain

is rapidly conveyed to auditory cortex and other parts of

sensory cortex. In the ventral midbrain, dopaminergic neu-

rons have been implicated in encoding motivational value

(seeking goals, evaluating outcomes, and value learning) or

motivational salience (orienting, cognitive processing, and

drive), thus providing an alerting signal (Bromberg-Martin

et al. 2010; Schulz 2008). Glutamatergic neurons may also

carry such information because selective activation of these

cells promotes reward (Wang et al. 2013). These cells may

also contribute to locomotor responses (Birgner et al. 2010;

Hnasko et al. 2012). GABAergic neurons showed persistent

activity during the interval between a reward-predicting

odor and the subsequent reward (Cohen et al. 2012). Our

findings also suggest that the projection from the dopami-

nergic ventral midbrain contributes to our previous obser-

vations of firing in auditory cortex that represents reward-

related information, including reward expectancy errors

(Brosch et al. 2011a). Current results are also compatible

with the proposal by Seamans and Yang (2004) for the

prefrontal cortex that the early activation by stimulating the

dopaminergic ventral midbrain initiates the persistent

activity states of cortical networks that allow for control of

cognitive functions, such as holding information about

auditory stimuli in a working memory buffer (Huang et al.

2012) or more generally to inform cortical networks when

certain information should be represented (Brosch et al.

2011b). It should be noted that in addition to fast mesocor-

tical information transmission by amino acid neurotrans-

mitters, the dopaminergic ventral midbrain also affects

auditory cortex on a slower timescale by the monoaminergic

neurotransmitter dopamine that is released from terminals in

this cortical structure and taken up by different dopamine

receptor types and might serve different roles (Stark and

Scheich 1997; Bao et al. 2001, 2003; Schicknick et al. 2008,

2012; Weis et al. 2012).

Acknowledgments The authors thank Dr. Jorg Stadler for MRI

measurements and Drs. Eike Budinger, Nikolaos Aggelopoulos, and

Daniel Zaldivar for valuable suggestions on the manuscript. The work

was supported by the Deutsche Forschungsgemeinschaft (DFG, SFB

779), the federal state of Saxony-Anhalt and the European Regional

Development Fund (ERDF 2007–2013) and the Russian Science

Foundation (RSCF 14-28-00229).

Conflict of interest There are no conflicts of interest.

References

Amunts K, Morosan P, Hilbig H, Zilles K (2012) Auditory system:

cyto-, myelo-, and receptor architecture of the auditory cortex.

In: Mai JK, Paxinos G (eds) The human nervous system, 3rd edn.

Elsevier Academic Press, San Diego, pp 1257–1287

Arsenault JT, Nelissen K, Jarraya B, Vanduffel W (2013) Dopami-

nergic reward signals selectively decrease fMRI activity in

primate visual cortex. Neuron 77(6):1174–1186

Atzori M, Kanold PO, Pineda JC, Flores-Hernandez J, Paz RD (2005)

Dopamine prevents muscarinic-induced decrease of glutamate

release in the auditory cortex. Neuroscience 134(4):1153–1165

Au-Young SM, Shen H, Yang CR (1999) Medial prefrontal cortical

output neurons to the ventraltegmental area (VTA) and their

responses to burst-patterned stimulation of the VTA: neuroana-

tomical and invivo electrophysiological analyses. Synapse

34:245–255

Bao S, Chan VT, Merzenich MM (2001) Cortical remodelling

induced by activity of ventral tegmental dopamine neurons.

Nature 412:79–83

Bao S, Chan VT, Zhang LI, Merzenich MM (2003) Suppression of

cortical representation through backward conditioning. Proc Natl

Acad Sci USA 100:1405–1408

Berger B, Gaspar P, Verney C (1991) Dopaminergic innervation of

the cerebral cortex: unexpected differences between rodents and

primates. Trends Neurosci 14(1):21–27

Bernardi G, Cherubini E, Marciani MG, Mercuri N, Stanzione P

(1982) Responses of intracellularly recorded cortical neurons to

the iontophoretic application of dopamine. Brain Res

245:267–274

Birgner C, Nordenankar K, Lundblad M, Mendez JA, Smith C, le

Greves M, Galter D, Olson L, Fredriksson A, Trudeau LE,

Kullander K, Wallen-Mackenzie A (2010) VGLUT2 in dopa-

mine neurons is required for psychostimulant-induced behavioral

activation. Proc Natl Acad Sci USA 107(1):389–394

Bjorklund A, Lindvall O (1984) Dopamine-containing systems in the

CNS. In: Bjorklund A, Hokfelt T (eds) Handbook of chemical

neuroanatomy: classical transmitter in the rat, vol 2. Elsevier/

North Holland, Amsterdam, pp 55–122

Bromberg-Martin ES, Matsumoto M, Hikosaka O (2010) Dopamine

in motivational control: rewarding, aversive, and alerting.

Neuron 68(5):815–834

Brosch M, Schulz A, Scheich H (1999) Processing of sound

sequences in macaque auditory cortex: response enhancement.

J Neurophysiol 82(3):1542–1559

Brosch M, Scheich H (2008) Tone-sequence analysis in the auditory

cortex of awake macaquemonkeys. Exp Brain Res

184(3):349–361

Brosch M, Selezneva E, Scheich H (2011a) Representation of reward

feedback in primate auditory cortex. Front Syst Neurosci 5:5

Brosch M, Selezneva E, Scheich H (2011b) Formation of associations

in auditory cortex by slow changes of tonic firing. Hear Res

271(1–2):66–73

Budinger E, Laszcz A, Lison H, Scheich H, Ohl FW (2008) Non-

sensory cortical and subcortical connections of the primary

auditory cortex in Mongolian gerbils: bottom-up and top-down

processing of neuronal information via field AI. Brain Res

1220:2–32

Campbell MJ, Lewis DA, Foote SL, Morrison JH (1987) Distribution

of choline acetyltransferase-, serotonin-, dopamine-beta-hydro-

xylase-, tyrosine hydroxylase-immunoreactive fibers in monkey

primary auditory cortex. J Comp Neurol 261(2):209–220

Brain Struct Funct

123

Carr DB, Sesack SR (2000) GABA-containing neurons in the rat

ventral tegmental area project to the prefrontal cortex. Synapse

38(2):114–123

Chiodo LA (1988) Dopamine-containing neurons in the mammalian

central nervous system: electrophysiology and pharmacology.

Neurosci Biobehav Rev 12(1):49–91

Chudasama Y, Robbins TW (2004) Dopaminergic modulation of

visual attention and working memory in the rodent prefrontal

cortex. Neuropsychopharmacology 29(9):1628–1636

Cohen JY, Haesler S, Vong L, Lowell BB, Uchida N (2012) Neuron-

type-specific signals for reward and punishment in the ventral

tegmental area. Nature 482(7383):85–88

Delgado JM (1965) Sequential behavior induced repeatedly by

stimulation of the red nucleus in free monkeys. Science

148(3675):1361–1363

Deniau JM, Thierry AM, Feger J (1980) Electrophysiological

identification of mesencephalic ventromedial tegmental (VMT)

neurons projecting to the frontal cortex, septum and nucleus

accumbens. Brain Res 189:315–326

Diana M, Garcia-Munoz M, Richards J, Freed CR (1989) Electrophys-

iological analysis of dopamine cells from the substantia nigra pars

compacta of circling rats. Exp Brain Res 74(3):625–630

Durstewitz D, Kroner S, Gunturkun O (1999) The dopaminergic

innervation of the avian telencephalon. Prog Neurobiol

59(2):161–195

Ferron A, Thierry AM, Le Douarin C, Glowinski J (1984) Inhibitory

influence of the mesocortical dopaminergic system on sponta-

neous activity or excitatory response induced from the thalamic

mediodorsal nucleus in the rat medial prefrontal cortex. Brain

Res 302(2):257–265

Fields HL, Hjelmstad GO, Margolis EB, Nicola SM (2007) Ventral

tegmental area neurons in learned appetitive behavior and

positive reinforcement. Annu Rev Neurosci 30:289–316

Freeman AS, Bunney BS (1987) Activity of A9 and A10 dopami-

nergic neurons in unrestrained rats: further characterization and

effects of apomorphine and cholecystokinin. Brain Res

405(1):46–55

Freeman AS, Meltzer LT, Bunney BS (1985) Firing properties of

substantia nigra dopaminergic neurons in freely moving rats.

Life Sci 36(20):1983–1994

Fritz J, Mishkin M, Saunders RC (2005) In search of an auditory

engram. Proc Natl Acad Sci USA 102(26):9359–9364

Fujisawa S, Buzsaki G (2011) A 4 Hz oscillation adaptively

synchronizes prefrontal, VTA, and hippocampal activities.

Neuron 72(1):153–165

Gao M, Liu CL, Yang S, Jin GZ, Bunney BS, Shi WX (2007)

Functional coupling between the prefrontal cortex and dopamine

neurons in the ventral tegmental area. J Neurosci 27:5414–5421

Gaspar P, Stepniewska I, Kaas JH (1992) Topography and collater-

alization of the dopaminergic projections to motor and lateral

prefrontal cortex in owl monkeys. J Comp Neurol 325(1):1–21

Gibson AR, Houk JC, Kohlerman NJ (1985) Magnocellular red

nucleus activity during different types of limb movement in the

macaque monkey. J Physiol 358:527–549

Gittelman JX, Perkel DJ, Portfors CV (2013) Dopamine modulates

auditory responses in the inferior colliculus in a heterogeneous

manner. J Assoc Res Otolaryngol 14(5):719–729

Grace AA (1991) Regulation of spontaneous activity and oscillatory

spike firing in rat midbrain dopamine neurons recorded in vitro.

Synapse 7(3):221–234

Grace AA, Bunney BS (1980) Nigral dopamine neurons: intracellular

recording and identification with L-dopa injection and histoflu-

orescence. Science 210(4470):654–656

Grace AA, Bunney BS (1983) Intracellular and extracellular electro-

physiology of nigral dopaminergic neurons—1. Identification

and characterization. Neuroscience 10(2):301–315

Grace AA, Bunney BS (1984a) The control of firing pattern in nigral

dopamine neurons: burst firing. J Neurosci 4(11):2877–2890

Grace AA, Bunney BS (1984b) The control of firing pattern in nigral

dopamine neurons: single spike firing. J Neurosci

4(11):2866–2876

Grace AA, Onn SP (1989) Morphology and electrophysiological

properties of immunocytochemically identified rat dopamine

neurons recorded in vitro. J Neurosci 9(10):3463–3481

Guyenet PG, Aghajanian GK (1978) Antidromic identification of

dopaminergic and other output neurons of the rat substantia

nigra. Brain Res 150(1):69–84

Happel MF, Jeschke M, Ohl FW (2010) Spectral integration in

primary auditory cortex attributable to temporally precise

convergence of thalamocortical and intracortical input. J Neuro-

sci 30(33):11114–11127

Happel MF, Deliano M, Handschuh J, Ohl FW (2014) Dopamine-

modulated recurrent corticoefferent feedback in primary sensory

cortex promotes detection of behaviorally relevant stimuli.

J Neurosci 34(4):1234–1247

Histed MH, Bonin V, Reid RC (2009) Direct activation of sparse,

distributed populations of cortical neurons by electrical micr-

ostimulation. Neuron 63(4):508–522

Hnasko TS, Hjelmstad GO, Fields HL, Edwards RH (2012)

Ventral tegmental area glutamate neurons: electrophysiolog-

ical properties and projections. J Neurosci 32(43):

15076–15085

Hosp JA, Pekanovic A, Rioult-Pedotti MS, Luft AR (2011) Dopa-

minergic projections from midbrain to primary motor cortex

mediate motor skill learning. J Neurosci 31:2481–2487

Houk JC, Gibson AR, Harvey CF, Kennedy PR, van Kan PL (1988)

Activity of primate magnocellular red nucleus related to hand

and finger movements. Behav Brain Res 28(1–2):201–206

Huang Y, Zacharias N, Konig R, Brosch M, Heil P (2012)

Physiological mechanisms of working memory in the auditory

cortex of humans and nonhuman primates. 34th ARO Meeting

Abstr 470:164

Hur EE, Zaborszky L (2005) VGLUT2 afferents to the medial

prefrontal and primary somatosensory cortices: a combined

retrograde tracing in situ hybridization study. J Comp Neurol

483(3):351–373

Irvine DR (1980) Acoustic input to neurons in feline red nucleus.

Brain Res 200(1):169–173

Jacob SN, Ott T, Nieder A (2013) Dopamine regulates two classes of

primate prefrontal neurons that represent sensory signals.

J Neurosci 33(34):13724–13734

Jay TM, Glowinski J, Thierry AM (1995) Inhibition of hippocampo-

prefrontal cortex excitatory responses by the mesocortical DA

system. NeuroReport 6(14):1845–1848

Jervis BW, Nichols MJ, Johnson TE, Allen E, Hudson NR (1983) A

fundamental investigation of the composition of auditory evoked

potentials. IEEE Trans Biomed Eng 30(1):43–50

Kajikawa Y, Schroeder CE (2011) How local is the local field

potential? Neuron 72(5):847–858

Kaur S, Rose HJ, Lazar R, Liang K, Metherate R (2005) Spectral

integration in primary auditory cortex: laminar processing of

afferent input, in vivo and in vitro. Neuroscience

134(3):1033–1045

Kennedy PR, Gibson AR, Houk JC (1986) Functional and anatomic

differentiation between parvicellular and magnocellular regions

of red nucleus in the monkey. Brain Res 364(1):124–136

Kiyatkin EA (1988) Morphine-induced modification of the functional

properties of ventral tegmental area neurons in conscious rat. Int

J Neurosci 41(1–2):57–70

Kiyatkin EA, Zhukov VN (1988) Impulse activity of mesencephalic

neurons on nociceptive stimulation in awake rats. Neurosci

Behav Physiol 18(5):393–400

Brain Struct Funct

123

Kudoh M, Shibuki K (2006) Sound sequence discrimination learning

motivated by reward requires dopaminergic D2 receptor activa-

tion in the rat auditory cortex. Learn Mem 13:690–698

Larsen KD, Yumiya H (1980) The red nucleus of the monkey.

Topographic localization of somatosensory input and motor

output. Exp Brain Res 40(4):393–404

Lavin A, Nogueira L, Lapish CC, Wightman RM, Phillips PE,

Seamans JK (2005) Mesocortical dopamine neurons operate in

distinct temporal domains using multimodal signaling. J Neurosci

25(20):5013–5023

Lewis DA, Campbell MJ, Foote SL, Goldstein M, Morrison JH

(1987) The distribution of tyrosine hydroxylase-immunoreactive

fibers in primate neocortex is widespread but regionally specific.

J Neurosci 7(1):279–290

Lidow MS, Goldman-Rakic PS, Gallager DW, Rakic P (1991)

Distribution of dopaminergic receptors in the primate cerebral

cortex: quantitative autoradiographic analysis using [3H]raclo-

pride, [3H]spiperone and [3H]SCH23390. Neuroscience

40(3):657–671

Lovell JM, Mylius J, Scheich H, Brosch M (2014) Hearing in action;

auditory properties of neurones in the red nucleus of alert

primates. Front Neurosci 8:105

Mantz J, Milla C, Glowinski J, Thierry AM (1988) Differential effects

of ascending neurons containing dopamine and noradrenaline in

the control of spontaneous activity and of evoked responses in

the rat prefrontal cortex. Neuroscience 27(2):517–526

Margolis EB, Lock H, Hjelmstad GO, Fields HL (2006) The ventral

tegmental area revisited: is there an electrophysiological marker

for dopaminergic neurons? J Physiol 577(Pt 3):907–924

Martin RF, Bowden DM (1996) A stereotaxic template atlas of the

macaque brain for digital imaging and quantitative neuroanat-

omy. Neuroimage 4(2):119–150

Massion J, Albe-Fessard D (1963) Duality of afferent sensory tracts

controlling the activity of the red nucleus. Electroencephalogr

Clin Neurophysiol 15:435–454

Matsumura M, Sawaguchi T, Kubota K (1990) Modulation of

neuronal activities by iontophoretically applied catecholamines

and acetylcholine in the primate motor cortex during a visual

reaction-time task. Neurosci Res 8(2):138–145

Mercuri N, Calabresi P, Stanzione P, Bernardi G (1985) Electrical

stimulation of mesencephalic cell groups (A9–A10) produces

monosynaptic excitatory potentials in rat frontal cortex. Brain

Res 338:192–195

Mitzdorf U (1985) Current source-density method and application in

cat cerebral cortex: investigation of evoked potentials and EEG

phenomena. Physiol Rev 65(1):37–100

Muller-Preuss P, Mitzdorf U (1984) Functional anatomy of the

inferior colliculus and the auditory cortex: current source density

analyses of click-evoked potentials. Hear Res 16(2):133–142

Nicholson C, Freeman JA (1975) Theory of current source-density

analysis and determination of conductivity tensor for anuran

cerebellum. J Neurophysiol 38(2):356–368

Oades RD, Halliday GM (1987) Ventral tegmental (A10) system:

neurobiology. 1. Anatomy and connectivity. Brain Res

434(2):117–165

Pirot S, Godbout R, Mantz J, Tassin JP, Glowinski J, Thierry AM

(1992) Inhibitory effects of ventral tegmental area stimulation on

the activity of prefrontal cortical neurons: evidence for the

involvement of both dopaminergic and GABAergic components.

Neuroscience 49:857–865

Pistis M, Porcu G, Melis M, Diana M, Gessa GL (2001) Effects of

cannabinoids on prefrontal neuronal responses to ventral

tegmental area stimulation. Eur J Neurosci 14(1):96–102

Ranck JB Jr (1975) Which elements are excited in electrical

stimulation of mammalian central nervous system: a review.

Brain Res 98(3):417–440

Sawaguchi T, Goldman-Rakic PS (1991) D1 dopamine receptors in

prefrontal cortex: involvement in working memory. Science

251(4996):947–950

Sawaguchi T, Goldman-Rakic PS (1994) The role of D1-dopamine

receptor in working memory: local injections of dopamine

antagonists into the prefrontal cortex of rhesus monkeys

performing an oculomotor delayed-response task. J Neurophysiol

71:515–528

Sawaguchi T, Matsumura M (1985) Laminar distributions of neurons

sensitive to acetylcholine, noradrenaline and dopamine in the

dorsolateral prefrontal cortex of the monkey. Neurosci Res

2(4):255–273

Sawaguchi T, Matsumura M, Kubota K (1986) Catecholamine

sensitivities of motor cortical neurons of the monkey. Neurosci

Lett 66(2):135–140

Sayers BM, Beagley HA, Henshall WR (1974) The mechanism of

auditory evoked EEG responses. Nature 247(5441):481–483

Scheibner T, Tork I (1987) Ventromedial mesencephalic tegmental

(VMT) projections to ten functionally different cortical areas in

the cat: topography and quantitative analysis. J Comp Neurol

259(2):247–265

Schicknick H, Schott BH, Budinger E, Smalla KH, Riedel A,

Seidenbecher CI, Scheich H, Gundelfinger ED, Tischmeyer W

(2008) Dopaminergic modulation of auditory cortex-dependent

memory consolidation through mTOR. Cereb Cortex

18:2646–2658

Schicknick H, Reichenbach N, Smalla KH, Scheich H, Gundelfinger

ED, Tischmeyer W (2012) Dopamine modulates memory

consolidation of discrimination learning in the auditory cortex.

Eur J Neurosci 35(763–74):2012

Schultz W (1998) Predictive reward signal of dopamine neurons.

J Neurophysiol 80:1–27

Schultz W, Dayan P, Montague PR (1997) A neural substrate of

prediction and reward. Science 275:1593–1599

Schultz W, Preuschoff K, Camerer C, Hsu M, Fiorillo CD, Tobler PN,

Bossaerts P (2008) Explicit neural signals reflecting reward

uncertainty. Philos Trans R Soc Lond B Biol Sci 363(1511):

3801–3811

Seamans JK, Yang CR (2004) The principal features and mechanisms

of dopamine modulation in the prefrontal cortex. Prog Neurobiol

74:1–58

Smith Y, Wichmann T, Delong MR (2013) Corticostriatal and

mesocortical dopamine systems: do species differences matter?

Nat Rev Neurosci 15(1):63

Stark H, Scheich H (1997) Dopaminergic and serotonergic neuro-

transmission systems are differentially involved in auditory

cortex learning: a long-term microdialysis study of metabolites.

J Neurochem 68:691–697

Steffensen SC, Svingos AL, Pickel VM, Henriksen SJ (1998)

Electrophysiological characterization of GABAergic neurons in

the ventral tegmental area. J Neurosci 18(19):8003–8015

Steinschneider M, Tenke CE, Schroeder CE, Javitt DC, Simpson GV,

Arezzo JC, Vaughan HG Jr (1992) Cellular generators of the

cortical auditory evoked potential initial component. Electroen-

cephalogr Clin Neurophysiol 84(2):196–200

Steinschneider M, Reser DH, Fishman YI, Schroeder CE, Arezzo JC

(1998) Click train encoding in primary auditory cortex of the

awake monkey: evidence for two mechanisms subserving pitch

perception. J Acoust Soc Am 104(5):2935–2955

Super H, Roelfsema PR (2005) Chronic multiunit recordings in

behaving animals: advantages and limitations. Prog Brain Res

147:263–282

Swanson LW (1982) The projections of the ventral tegmental area

and adjacent regions: a combined fluorescent retrograde tracer

and immunofluorescence study in the rat. Brain Res Bull

9(1–6):321–353

Brain Struct Funct

123

Szabo J, Cowan WM (1984) A stereotaxic atlas of the brain of the

cynomolgus monkey (Macaca fascicularis). J Comp Neurol

222(2):265–300

Ter-Mikaelian M, Sanes DH, Semple MN (2007) Transformation of

temporal properties between auditory midbrain and cortex in the

awake Mongolian gerbil. J Neurosci 27(23):6091–6102

Thierry AM, Blanc G, Sobel A, Stinus L, Golwinski J (1973)

Dopaminergic terminals in the rat cortex. Science 182(4111):

499–501

Thierry AM, Deniau JM, Herve D, Chevalier G (1980) Electrophys-

iological evidence for non-dopaminergic mesocortical and

mesolimbic neurons in the rat. Brain Res 201(1):210–214

Thierry AM, Godbout R, Mantz J, Pirot S, Glowinski J (1992)

Differential influence of dopaminergic and noradrenergic affer-

ents on their target cells in the rat prefrontal cortex. Clin

Neuropharmacol 15 Suppl 1 Pt A:139–140

Tseng KY, Mallet N, Toreson KL, Le Moine C, Gonon F, O’Donnell

P (2006) Excitatory response of prefrontal cortical fast-spiking

interneurons to ventral tegmental area stimulation in vivo.

Synapse 59(7):412–417

Wang HL, Qi J, Morales M (2013) Activation of ventral tegmental

area glutamate neurons is rewarding. SfN Abstr 389:05

Watanabe Y, Kajiwara R, Takashima I (2009) Optical imaging of rat

prefrontal neuronal activity evoked by stimulation of the ventral

tegmental area. NeuroReport 20:875–880

Weis T, Puschmann S, Brechmann A, Thiel CM (2012) Effects of L-

dopa during auditory instrumental learning in humans. PLoS

ONE 7(12):e52504

Williams SM, Goldman-Rakic PS (1998) Widespread origin of the

primate mesofrontal dopamine system. Cereb Cortex 8:321–345

Yamaguchi T, Wang HL, Li X, Ng TH, Morales M (2011)

Mesocorticolimbic glutamatergic pathway. J Neurosci

31(23):8476–8490

Zilles K, Palomero-Gallagher N, Grefkes C, Scheperjans F, Boy C,

Amunts K, Schleicher A (2002) Architectonics of the human

cerebral cortex and transmitter receptor fingerprints: reconciling

functional neuroanatomy and neurochemistry. Eur Neuropsy-

chopharmacol 12(6):587–599

Brain Struct Funct

123

View publication statsView publication stats